A team of researchers from the University of Virginia Cancer Center has developed a new method to identify high-risk patients with acute myeloid leukemia (AML). This method could lead to more personalized treatments and better patient outcomes. Traditionally, doctors use gene and chromosome analysis to identify high-risk patients, but this method has its limitations. The team found that measuring specific molecules in cancer cells can help identify patients at risk of poor outcomes.
The research team, led by B. Bishal Paudel, PhD, used machine learning to analyze molecules called sphingolipids in cancer cells. These molecules are thought to play key roles in AML development and treatment resistance. By measuring these lipids, the team could classify AML into two subtypes. Patients in the high-risk subtype had double the risk of treatment failure compared to those in the other subtype.
Kevin A. Janes, PhD, another researcher from the UVA Cancer Center, noted that while lipids are harder to measure than genes, the study suggests it’s worth the effort. The team used the data to understand the relationship between genes and sphingolipid subtype. This showed how subtypes could be accurately identified in more patients than those measured directly.